From "Traffic Apocalypse/Google’s AI Overviews are killing click-throughs to news sites" (Columbia Journalism Review).
July 31, 2025
"Worldwide search traffic has fallen by 15 percent in the past year.... Now that AI-generated summaries are being integrated into search results..."
"... anyone looking for information has less reason to click through to the websites where that information originates. For media publishers whose business models rely on referral traffic to bring them advertising revenue, this shift feels nothing short of catastrophic.
There’s no getting around the decline in traffic. Last week, the Pew Research Center released a report showing not only that people who see an AI-generated summary on Google search are significantly less likely to click on external links than users who don’t, but that people almost never click on the links included in the AI summary. (They do so just 1 percent of the time.).... One strategy is trying to demand compensation from AI companies for crawling their content. Media companies are also investing in their own channels that deliver content directly to readers. They are launching new subscriptions, newsletters, events, membership programs, and even platforms and apps.
Wired’s Katie Drummond wrote recently that the... trick... is to 'connect our humans to all of you humans.'...."
Posts
50 comments:
I've stopped using Google in favor of xAI for most queries. The reason is that journalists and media organizations in the United States are absolute fucking assholes. I, and 78% of other Americans want them all to die slow deaths in a vat of acid.
Go fuck yourselves.
AI is basically just stealing the stuff. It's a giant plagiarism machine...and that's its good side. Change my mind.
I...want them all to die slow deaths in a vat of acid.
You know what would be better? If they all had to learn a trade and do tangible work to feed themselves. Could you imagine the wailing and gnashing if they had to go out and paint houses or do landscaping?
Eventually, tho, they might actually develop some self-respect.
I often find that these pop-up AI summaries are incomplete, misleading, or sometimes even false. If I'm doing anything more that a cursory search, I always click on several of the links to better understand what I'm looking for.
people almost never click on the links included in the AI summary
This is surprising as the lack of included links makes the AI result much less useful. I hate having to go search for the AI's sources to confirm or flesh out what it says.
"but that people almost never click on the links included in the AI summary. (They do so just 1 percent of the time.)."
I guess I'm an oddball. I click through frequently to the source because I don't trust AI to give correct information. This is linked to once reading an AI summary regarding calisthenics that can be done at home to strengthen back muscles that started out by saying make sure you were lying on your stomach flat on the floor and then secondly that the bottom of your feet were also flat on the floor.
AI summaries are great. Unless you actually want to be right. In which case, you have to check through more than one AI response to your question. Then, just to be sure, check an additional source, either a book (remember those?) or your old Google Search to get you to a source that is or was recently human.
Aside from that AI is cute and has good genes, if I can say that.
The AI overviews are a hassle because they are so often wrong and off topic.
“I often find that these pop-up AI summaries are incomplete, misleading, or sometimes even false.”
Mirrors your comments, except they are usually false.
Also killing Google traffic, the fact of terminal convergence making their search engine almost unusable unless you explicitly know what you're looking for.
Zuck's gonna kill that golden goose if he ain't careful.
People who value Content will pay for valuable content.
But people who have boring opinions will have their stuff summarized by AI.
Sure, blame it on AI.
Re Google I am enjoying its new toy cannibalizing its top revenue generator and replacing those earnings with news releases instead of profits.
Shouldn't been evil!
Even if you click a few links from the summary, the time saved not trying to get ads to go away from old school basic searches is a huge stress and time saver. Plus, AI has winnowed down the links so they’re usually more reactive to what you’re actually asking. Or attempting to ask, since I often need to tune my query…. I’m only human.
Seems like an equilibrium will eventually be reached. If content generators aren't getting eyeballs, they will cut back or go out of business. Which will give the AI less to work with. Which means it will get fewer eyeballs. Some new equilibrium states could include: the AI company sharing more money with the content generators; the AI company buying and operating the content generators; as the AI gets better, generating primary content with AI (eg, an AI "reporter" can rephrase a govt press release as well or better than a human "reporter"); or, as the cost of AI drops, the content generators doing their own AI summaries, of their own and others' content.
RR
JSM
I suppose it's only a matter of time before ads begin appearing in AI responses. Makes sense, no?
connect our humans to all of you humans
The problem here is that their humans suck. They are mindless drones spouting DNC talking points, nothing more.
And people are no longer willing to pay for that (as opposed to Trump's first term, when they were).
Couldn't happen to a more deserving group of scum
Bob Boyd said...
AI is basically just stealing the stuff. It's a giant plagiarism machine...and that's its good side. Change my mind.
Well, it's robbing people who I want dead. So that makes it a feature, not a bug
Am I the only one passing over the AI summary and looking for the actual links myself? I don't trust them for accuracy.
I get AI suggestions all the time... and I ignore them when I google.. that is all you have to do folks.. ignore it. Cause most of the time the AI is just so general... and I need specific info.
AI is embedded in our devices and apps. So it's reading our emails and private texts. The implications are massive.
AI’s misleading, incomplete, and false info is free. The NYT isn’t.
I use the AI-generated summary sparingly and I never, never, never use Google Search. I use DuckDuckGo instead.
Grok is pretty damn helpful but I never trust it implicitly.
I am far more suspicious of the search engines themselves, and the way they massage their results (especially on the first page) to give advantage to the Democrat Party and reward their advertisers.
I've gotten pretty good at tailoring my search requests to evade the leftist censorship. Sometimes including terms like "disinformation" and following the first link to the official propaganda mouthpiece is a good way to be told about the samizdat outlets where the truth is actually to be found. It's a bit like the way the Russians had to read Pravda in Soviet days.
And speaking of Russia, the Yandex search engine is much like the old uncensored Google circa 2010, for anything not touching on Russian topics. I wouldn't trust it to give me truthful links about Putin or the Ukraine invasion, but it's very good for information that is in bad odor with the leftist would-be tyrants trying to control so-called disinformation in the West.
Such subterfuges will have to do until Nina Jankowicz and her henchmen go to Federal prison for decades for violating our First Amendment rights, and our British cousins wake up and nail Sir Keir's severed head to the Tower Bridge to feed the ravens.
If you like to follow the trail and explore and aren't satisfied with a condensation, it's a little frustrating. If you're questioning the official view, you don't need to be confronted with it again before the digging begins. Sometimes you may want to know if other people make the same connections you do, rather than hear what the authorities want you to think.
Online you may see scores of clickbait articles. You may want to know who the five performers hated by Kevin Costner are, or why you should put aluminum foil around your door knobs, without going to the clickbait site. So you turn to AI (and it often turns out that there's absolutely nothing behind the stories). This would tend to put the clickbait sites out of business (the same problem the legacy media will be having). Can they declare their material copyrighted and out of bounds to AI scrapers and scavengers? Can they let Google list them in results and still deny Google AI the right to summarize the entire article?
I type in direct links to the information I seek. I don't trust generic search engines. 😜
"connect our humans to all of you humans"
They actually think we want to be friends with them? The gall. The absolute gall! How disconnected from reality are these morons?
The only way I want to connect with a member of the media is through an uppercut. Journalists in the United States are sociopaths and many outright psychopaths. We want NOTHING to do with ANY of you broken people.
Those coders better learn to mine coal.
If the clickbait providers weren't so pervasive and obnoxious, I might click every once in a while. As it is, AI is generally the "first pass"; then I proceed carefully. About 20 percent of the time the AI information is clearly wrong. The political biases of the AI services are generally indistinguishable from those of the liberal media.
The internet is eating the media. First it was newspapers, then television, now it’s eating itself.
Google brought this on itself by turning a once decent search engine into garbage site that promotes the same handful of links over and over. I hate clicking through on Google. They sold out for ad revenue and political advocacy.
Several weeks ago I was trying to find a basic statement or summary of the Republican congress's position regarding Medicaid "cuts" in the BBB.
I had to go 4 pages into Google before I could find anything that wasn't Democrat/Healthcare corporation propaganda--i.e., anything that actually stated that the "cuts" were to eliminate deceased people, illegal non-citizens, and healthy people who refuse to work. Google is powerful and fast--and absolutely biased.
https://brusselsmorning.com/ais-narcissistic-takeover-and-end-of-reality-the-narcissist-as-artificial-intelligence/54887/
Why look elsewhere when the Great and Benevolent Grok has spoken?
WAAAAAY back in 1975 I had an engineering professor who said he wasn't as concerned about Garbage In - Garbage Out as he was garbage in GOSPEL out.
Right again, Professor Lee,
asking for "[town name, state] hourly" gets you an ai forecast for a random day made at a random time for that town. AI hasn't figured out to say "at time x the forecast is," not to mention the hallucination problem. What is the usual forecast for this town, i.e.
Web search for how to load a file over usb to an android tablet have so far failed. I suspect, unlikely though it is, I just got two consecutive android tablets, as the battery seems unable to charge on either of them in spite of messages saying that the battery is charging rapidly. Anyway Amazon took them back with no problem.
Also tablets seem unable to connect to a nonbroadcasting wifi even if you tell them its name.
Still some bugs in the design.
I don't care what anyone says, AI is going to increase the monetary value of real writers.
Sometimes including terms like "disinformation" and following the first link to the official propaganda mouthpiece is a good way to be told about the samizdat outlets where the truth is actually to be found.
+false, +baseless, +debunked, +"without evidence", +scurrilous, and +discredited also work for this purpose. It's a nice bit of aikido to get past the left-wing censors.
Didn't several countries in recent years pass laws that required Google (and other search engines) to pay when they show links to news sites? They treid to make the case that search engines were stealing revenue from the news sites by showing thumbnails and quoting text from news sites. The governments said that they were going to take the royalty money that they were going to get from search engines and give it directly to the news outlets. I believe that Google's response was to stop showing links to news sites in those and a lot of people felt that Google doing this would actually cost the news sites more money because now they wouldn't be able to take advantage of the click throughs to their news articles.
If Google has to pay for this, it makes sense that they would tweak their search engines to avoid this expense.
https://www.aljazeera.com/economy/2022/5/11/google-to-pay-over-300-european-news-publishers-for-content
boatbuilder said...
"Several weeks ago I was trying to find a basic statement or summary of the Republican congress's position regarding Medicaid "cuts" in the BBB.
I had to go 4 pages into Google before I could find anything that wasn't Democrat/Healthcare corporation propaganda--i.e., anything that actually stated that the "cuts" were to eliminate deceased people, illegal non-citizens, and healthy people who refuse to work. Google is powerful and fast--and absolutely biased."
I skirted that by adding "Congressional Record" to the search argument, and was taken to the summary of all bill provisions at Congress.gov. Each summary references the cite in the USC where the changes took place.
https://www.congress.gov/index.php/bill/119th-congress/house-bill/1
For absolute basic information it’s fine. But as others have noted, invariably wrong. The links at the end of statements often do not contain anything like the statement made.
Also I’ve noticed the ‘AI’ summary taking up more room on the screen, the following links less relevant - I believe this is deliberate.
I use AI for very specific technical tasks that would require a dozen searches and sub-searches. It's gets to the actual answer when your question is detailed and specific. Like how to operate particular machine and if it's compatible with something else, or if it can handle a specific task. Problem is it gets it wrong a lot.
I had Grok program a CNC g-code job for me, and although it provided some useful stuff, it was far from the best way to do it, and it would never actually run.
It was very helpful in finding me an audio amplifier that was compatible with both my receiver and my speakers on my pontoon boat, and it worked out great. The boat is rocking now at 3 times the volume and clarity. I'm the Commodore of highest regard among all who sail with me thanks to my excellent tunes.
Bring back Alta Vista!
Ask it "Who killed JFK?" and see what you get.
People need information. Not so long ago they relied on print. Then Google. Now AI. The status quo will always be impermanent.
Seems to me that one of the purposes of AI search is reducing the number of searches you need to do to find an answer, so it makes sense there would be less searching.
Politics aside, if you want to search for something very specific the AI summary can be helpful. For example, I was trying to replace a circular saw blade, and wanted to know if I should turn the allen wrench screw clockwise or counter clockwise to remove the old blade. I don't need to wade through 15 sites (and get exposed to ads taking up screen real estate trying to sell me cufflinks, or something equally useless) to find the info (don't get me started about getting links to videos.) So I'd rather see if a summary will answer my question, or at least a good link to a step by step instruction.
It probably won't last, I remember when Google was great, for example, but AI search is just completely superior. I don't use the summary, I just search from ChatGPT. Grok is probably just as good.
Post a Comment
Please use the comments forum to respond to the post. Don't fight with each other. Be substantive... or interesting... or funny. Comments should go up immediately... unless you're commenting on a post older than 2 days. Then you have to wait for us to moderate you through. It's also possible to get shunted into spam by the machine. We try to keep an eye on that and release the miscaught good stuff. We do delete some comments, but not for viewpoint... for bad faith.